The aims of this study were to evaluate whether treatment with tumor necrosis factor (TNF) inhibitors in patients with rheumatoid arthritis (RA) affects the risk of developing severe extraarticular rheumatoid arthritis (ExRA) manifestations and to investigate potential predictors for developing ExRA. A dynamic community-based cohort of patients with RA was studied (n = 1977). Clinical records were reviewed and cases of severe ExRA were identified. Information on exposure to TNF inhibitors was obtained from a regional register. Exposure to TNF inhibitors was analyzed in a time-dependent fashion and the incidence of severe ExRA in exposed patients was compared with the incidence in unexposed patients. Cox regression models were used to assess potential predictors of severe ExRA. During treatment with TNF inhibitors, there were 17 patients with new onset of severe ExRA in 2400 person-years at risk (PY; 0.71/100 PY, 95% CI 0.41-1.13) compared with 104 in 15,599 PY (0.67/100 PY, 95% CI 0.54-0.81) in patients without TNF inhibitors. This corresponded to an incidence rate ratio of 1.06 (95% CI 0.60-1.78). The age- and sex-adjusted HR for ExRA in anti-TNF-treated patients was 1.21 (95% CI 1.02-1.43), with similar findings in models adjusted for time-dependent Health Assessment Questionnaire and propensity for anti-TNF treatment. Male sex, positive rheumatoid factor (RF), long disease duration, and greater disability were predictors for ExRA. This study suggests that patients treated with TNF inhibitors are at a slightly increased risk of developing severe ExRA. RF-positive patients with disabling disease of long duration were more likely to develop severe ExRA.